Date of Award
Spring 2024
Document Type
Dissertation
Degree Name
Doctor of Philosophy (PhD)
Department
Economics
First Advisor
Arkolakis, Costas
Abstract
The first chapter of this dissertation uses parcel-level tax assessment and sales transaction data to estimate the stringency of housing unit density restrictions at the census tract level for 35 metropolitan areas in the United States, focusing on the period between 2000 and 2020. I do this by inferring the marginal regulatory "cost" of building an additional housing unit, conditioning on a parcel's total acreage and developed floor space. Regression on variables from the Wharton land use survey offers basic validation of this measure. I show that the inferred unit regulatory costs vary drastically across cities and neighborhoods; they tend to increase in both distance from the central business district and median household income and decrease in existing population density and college-educated population share. Finally, I show that regulatory costs tend to be cyclical, as the market incentive for developers to build at higher densities is strongest during boom periods. The second chapter embeds density restrictions in an equilibrium model of housing markets where forward-looking landowner-developers choose the size and number of units to build, and households, differentiated by income, choose locations and units in which to live. I estimate the supply model using parcel-level housing data from CoreLogic and the demand model using household-level microdata from the Census and American Community Survey. I then use the estimated model to analyze a wide-reaching counterfactual reduction in density restrictiveness in greater Washington, D.C. beginning in the year 2000; the policy reduces regulatory costs to the 25th percentile level in tracts that are at least 25% undeveloped as of 2000. This leads to a 39% increase in the number of new units built between 2000 and 2020, a 4.2% decline in mean rent, and 3% welfare gains for low-income households (as a share of income) by 2020; higher-income households, especially renters, stand to lose from the policy due to a reduction in the number of large, low-density units. The change in policy also leads to more in-migration by lower-income households and attenuates income sorting in lower-income neighborhoods. The final chapter investigates the extent to which the geography of jobs drives occupational and spatial mobility in the United States. I first show that, between the 1990s and 2010s, metro areas and states generally became more similar to each other in terms of their occupational compositions, while occupations tended to converge in terms of their geographic compositions. I then turn to a highly stylized, but empirically flexible, model of occupation and location choice to quantify the role that spatial frictions play in occupational mobility, and vice versa. While spatial frictions have a non-negligible effect on switching into and out of particular occupations that are heavily concentrated in select rural or urban areas, the model does not attribute the overall decline in occupational mobility since the early 1990s to these frictions. Similarly, the contemporaneous decline in spatial mobility cannot be attributed to occupational re-skilling frictions.
Recommended Citation
Tokman, Anthony Elias, "Essays on Housing, Inequality, and Spatial Mobility" (2024). Yale Graduate School of Arts and Sciences Dissertations. 1462.
https://elischolar.library.yale.edu/gsas_dissertations/1462
